OR18

Tongue swabs are a promising sampling alternative to diagnose pulmonary tuberculosis (TB), especially in patients with difficulty to produce sputum. Molecular testing of tongue swabs showed encouraging sensitivity compared to sputum. Despite improvements in genotypic drug-susceptibility testing (DST), phenotypic DST remains the gold standard for many anti-TB drugs, requiring a cultured isolate. The elimination of commensal flora is critical to avoid culture contamination. In this study, we compared the performance of two simple decontamination methods on spiked tongue swabs taken from healthy volunteers.

The Kudoh-Ogawa (KO) method inactivates commensal flora using NaOH followed by inoculation directly on slightly acidic Ogawa, while decontamination by cetylpyridinium chloride (CPC) is followed by inoculation on neutral pH Löwenstein-Jensen (LJ). To evaluate whether Mycobacterium tuberculosis (Mtb) but not the commensal flora from tongue swabs can survive CPC/LJ and KO, we processed tongue swabs (Copan Floqswab) taken from 20 volunteers with both methods, versus control swabs without decontamination. Half of the samples were spiked with the H37Rv reference strain (10 µl of 10⁻⁴ dilution McFarland #1). Both methods reduced contamination significantly (85% to 0.5% for KO and 30% to 2.5% for CPC/LJ), while Mtb recovery was high (97%). Storage in CPC for less than 7 days did not impact Mtb growth, while recovery of H37Rv gradually decreased thereafter, with less abundant and slower growth.

Our results suggest that current sputum decontamination methods also work for tongue swabs. Work in progress assesses culture in liquid systems and evaluates swabs from pulmonary TB patients in field conditions.